HEALTH AI NETWORK
  • Home
  • Primer
  • Ventures
  • About

Technology

 Your latest on new learnings in AI technology and tools
Picture
May 2019

Leveraging implicit knowledge in neural networks for functional dissection and engineering of proteins

Check it out here


Picture
May 2019

Perspective: Stop explaining black box machine learning models for high stakes decisions and use interpretable models instead

Read it here


Picture
May 2019

Integrating IoT into healthcare, agriculture, and transportation use cases

Read more here


Picture
May 2019

Seamless data architecture to support artificial intelligence success in healthcare

See more here


Picture
April 2019

Google launches end-to-end AI platform

Read more here


Picture
April 2019

NVIDIA partners with American College of Radiology to bring free AI tools

See more here


Picture
April 2019

Speech synthesis from neural decoding of spoken sentences

Check it out here


Picture
April 2019

Predicting scheduled hospital attendance with AI

Read more here


Picture
April 2019

Applications of supervised and unsupervised machine learning in Pharmaceutical drug development

See more here


Picture
April 2019

Development and validation of a machine-learning risk algorithm for major complications and death after surgery

Read more here


Picture
March 2019

Who made that decision: you or an algorithm?

Read more here


Picture
March 2019

A computer vision system for deep learning-based detection of patient mobilization activities in the ICU

Check it out here


Picture
March 2019

Death by a Thousand Clicks: Where EHRs Went Wrong

Read more here


Picture
March 2019

New Coursera Course: AI For Everyone - a course for non-engineers

Check it out here


Picture
March 2019

Been Kim: Building a translator to understand how neural networks work

See more here


Picture

​​March 2019


Back to the Basics: Ten Commandments for Learning to Code

Read more here


Picture
March 2019

As AI establishes its foothold in the healthcare space, the industry looks to identify better options to aggregate data (the keystone of AI/ML)

See more here


Picture
March 2019

This recent paper shows that deep neural networks can ID histologic patterns on slides of resected lung adenocarcinoma

Learn more here


Picture
February 2019

Review: ​Machine Learning Over Logistic Regressions

Read more here


Picture
February 2019

Predictive Modeling of the Hospital Readmission Risk from Patients’ Claims Data Using Machine Learning: A Case Study on COPD
​
Read more here


Picture
February 2019

Using Data Science to Predict Social Determinants of Health

Learn more here or here


Picture
 February 2019

In a recent trial, the new computational tool, TEXlab, uses 4 biological characteristics (structure, shape, size, and genetics) over ovarian tumors to generate a Radiomic Prognostic Vector (RPV) score

Read more here or here


Picture
February 2019

SHAP (SHapley Additive exPlanations) -  a unified approach to interpreting model predictions

Learn more here


Picture
February 2019

Machine learning for sepsis identification

Read more here


Picture
February 2019

Medical AI Safety: Doing it Wrong. In healthcare, model performance does not always equate to outcomes

See more here


Picture
January 2019

From MMIC: 400K labeled chest radiographs for future AI projects

Learn more here


Picture
January 2019

Practical implementation of AI technologies in healthcare. Hint: it includes addressing data sharing and data standardization

See it here


Picture
January 2019

Guidelines for the use of reinforcement learning in healthcare

Learn more here


Picture
January 2019

Davos: Importance of Privacy and Trust in New Technologies

Read more here


Picture
January 2019

AI voted as the most disruptive technology in healthcare. Does this translate to most impactful?

Check it out here


Picture
December 2018

Technology trends in healthcare AI - predictions for 2019

See more here


Picture
December 2018

"A(P)Is": National Coordinator emphasizes the need for APIs in AI

See more here


Picture

December 2018

Clinical Data Support Systems in the Age of AI: Challenges to Adoption and Solutions

​Read more here


Picture
November 2018

Encoder-Decoder algorithm can help identify chief complaints from EHR data through natural language processing

See more here


Picture
November 2018

Generative Adversarial Networks (GANs) used to generate finger print scans that can fool phone locks

Learn more here


Picture
November 2018

What’s really scary about AI is not its smartness but its dumbness: what AI can and cannot do. 

Read about it here


Picture
November 2018

​AI relates often relates imperceptible observations to outcomes in a fashion that’s unapologetically oblivious to mechanism. This challenges physicians and drug developers by explicitly severing utility from foundational scientific understanding.

Not sure what that means? See more here


Picture
November 2018

​FHIR Server for Azure empowers developers with open source software to move clinical health data into the Microsoft Cloud with the emerging standard HL7 FHIR (Fast Healthcare Interoperability Resources).

Check it out here


Picture


​November 2018

JAMA Viewpoint emphasizes the importance of interpretability and workflow integration for predictive models in healthcare. The authors describe a framework for evaluating Clinical Decision Support Systems.

Learn more here


Picture

​October 2018

21st Century Cures Act Health IT Provisions Promotes Interoperability and Data Exchange

Learn more here


Picture
October 2018

Qure.ai developed a deep learning model to identify acute findings in head CTs. After training on 290,055 CTs, they achieved above 90% sensitivity and specificity for every acute finding they assessed on a testing set of 21,095 scans. They also demonstrated that their algorithm was as sensitive as a radiologist, albeit not as specific.

See more here


Picture
October 2018
​

Next generation Natural Language Processing from Google, vastly outperforming previous algorithms on industry benchmarks.

Learn more here


Picture

​October 2018

“Why Should I Trust You?” Explaining the Predictions of Any Classifier.
​

Check it out here


Picture

​October 2018

A Guide to ML for Difficult Projects: How to Deliver

Learn more here


Picture
October 2018

What are machine learning algorithms really picking up on in medical images? This study by Finlayson et al. shows that subtle changes to images that are imperceptible by humans can completely fool a computer vision algorithm. The authors also discuss why certain actors in the healthcare space may be incentivized to falsify data and fool these algorithms.

Read more here


Picture
October 2018

Kaggle is hosting the RSNA pneumonia detection challenged based on the NIH Chest X-ray data set comprising 112,120 X-ray images with disease labels from 30,805 unique patients. It is an exciting example of data scientists exploring clinical, open-source data sets.
​
Check it out here


Picture

October 2018

Google introduced a companion to Google Scholar called Google Dataset search. It will allow researchers and scientists to search through a wide variety of open source data sets. It is currently in Beta.

Learn more here


Picture
October 2018

Knowing ground truth is critical for training any supervised machine learning algorithm. However, in practice, we often can’t ensure that our training labels are perfect. Han et al. have generated “a meta algorithm called Pumpout to overcome the problem of memorizing noisy labels.”

Learn more here


Picture

September 2018

DARPA recently launched a new initiative -  Lifelong Learning Machines (L2M) programs - draws inspiration from biological systems and  seeks to develop fundamentally new ML approaches that allow systems to adapt continually to new circumstances without forgetting previous learning.
​

See the details here


Picture
September 2018

The AI Next Campaign is a HUGE, $2 billion project that DARPA is undertaking to bring about the so called “3rd generation of AI”. Some of that funding will go to computational biologists.
​

Check it out here


Picture
September 2018

What about my HIPPA data?! Amazon Web Services (AWS), an extremely popular cloud computing platform, is forging its way into Healthcare. Healthcare professionals can use AWS to process their HIPPA protected information with these guidelines.

Learn more here


Picture

​
​September 2018


Study finds that substantial differences exist across the common results of cardiac imaging modalities.

Read about it here


Picture



​September 2018

Sharing models instead of data: a potential remedy for medical data sharing problem?

Learn more here


Picture

September 2018

A recent study shows high accuracy and scalability of deep learning of electronic health records.

Details here


Interested in specific topics? Let us know at [email protected]
Powered by Create your own unique website with customizable templates.
  • Home
  • Primer
  • Ventures
  • About